2. Improving Machine Learning Models:-
• Businesses today depend on machine learning to
optimize and scale their operations.
• When using this analytical tool, you’ll be able to
generate essential data-driven insights.
• Creating a high-performance machine learning
(ML) model is quite challenging.
3. Continued……
• It’s even more challenging to boost the
performance of a machine learning model to
produce reliable and correct results.
• Data scientists acknowledge this as they often
face a hard time testing a model’s
performance to increase its accuracy.
4. Continued……
• If your ML model is struggling to deliver accurate
and reliable results, here are ten effective ways to
boost its performance.
• 1. Studying Learning Curves
• 2. Using Cross- Validation Correctly
• 3. Choosing the Right Error or Score Metric
• 4. Searching for the best Hyper-Parameters
• 5. Testing Multiple Models
5. Continued……
• 6. Averaging Models
• 7. Staking Models
• 8. Applying Feature Engineering
• 9. Selecting Features and Examples
• 10. Looking for More Data
6. Studying learning curves
• As a first step to improving your results, you
need to determine the problems with your
model.
• Learning curves require you to verify against a
test set as you vary the number of training
instances.
7. Continued……
• You’ll immediately notice whether you find
much difference between your in-sample and
out-of-sample errors.
• A wide initial difference is a sign of estimate
variance; conversely, having errors that are
both high and similar is a sign that you’re
working with a biased model.
8. Continued…….
• A learning model of a Machine Learning model
shows how the error in the prediction of a
Machine Learning model changes as the size of
the training set increases or decreases.
• Before we continue, we must first understand
what variance and bias mean in the Machine
Learning model.
9. • Bias:
• It is basically nothing but the difference between the
average prediction of a model and the correct value of
the prediction.
• Models with high bias make a lot of assumptions about
the training data.
• This leads to over-simplification of the model and may
cause a high error on both the training and testing sets.
• However, this also makes the model faster to learn and
easy to understand. Generally, linear model algorithms
like Linear Regression have a high bias.